Kernel-based linear system identification using augmented Krylov subspaces
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913104258924544 |
|---|---|
| author | Matti, Fabio Andersen, Martin Skovgaard Chen, Tianshi Kressner, Daniel |
| author_facet | Matti, Fabio Andersen, Martin Skovgaard Chen, Tianshi Kressner, Daniel |
| contents | We propose a novel Krylov subspace method for estimating the finite impulse response (FIR) of a one-dimensional linear time-invariant systems. The method approximates the system's FIR using a kernel-based formulation combined with hyperparameter selection based on maximum likelihood estimation (MLE), which requires repeated evaluation of two terms: The data fit $\boldsymbol{y}^{\top} (λ\boldsymbol{I} + \boldsymbol{A})^{-1} \boldsymbol{y}$ and the model complexity $\log(\det (λ\boldsymbol{I} + \boldsymbol{A}))$, where $\boldsymbol{A}$ is a certain positive semidefinite matrix that admits fast matrix--vector products and $λ> 0$ is a regularization parameter. Instead of approximating these two quantities separately, we jointly approximate them using a single augmented Krylov subspace for $\boldsymbol{A}$. One major benefit of augmentation is that we obtain accelerated convergence when approximating the data fit quadratic form, through implicit preconditioning. Thanks to the shift invariance of Krylov subspaces, the extracted approximations can be used to evaluate the MLE objective for many values of $λ$ at little additional cost. We derive error bounds for the approximations, reflecting the benefits of augmentation demonstrated through multiple numerical experiments. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_08362 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Kernel-based linear system identification using augmented Krylov subspaces Matti, Fabio Andersen, Martin Skovgaard Chen, Tianshi Kressner, Daniel Numerical Analysis 62C10, 65F08, 65F22, 65F40, 93B15, 93B30 We propose a novel Krylov subspace method for estimating the finite impulse response (FIR) of a one-dimensional linear time-invariant systems. The method approximates the system's FIR using a kernel-based formulation combined with hyperparameter selection based on maximum likelihood estimation (MLE), which requires repeated evaluation of two terms: The data fit $\boldsymbol{y}^{\top} (λ\boldsymbol{I} + \boldsymbol{A})^{-1} \boldsymbol{y}$ and the model complexity $\log(\det (λ\boldsymbol{I} + \boldsymbol{A}))$, where $\boldsymbol{A}$ is a certain positive semidefinite matrix that admits fast matrix--vector products and $λ> 0$ is a regularization parameter. Instead of approximating these two quantities separately, we jointly approximate them using a single augmented Krylov subspace for $\boldsymbol{A}$. One major benefit of augmentation is that we obtain accelerated convergence when approximating the data fit quadratic form, through implicit preconditioning. Thanks to the shift invariance of Krylov subspaces, the extracted approximations can be used to evaluate the MLE objective for many values of $λ$ at little additional cost. We derive error bounds for the approximations, reflecting the benefits of augmentation demonstrated through multiple numerical experiments. |
| title | Kernel-based linear system identification using augmented Krylov subspaces |
| topic | Numerical Analysis 62C10, 65F08, 65F22, 65F40, 93B15, 93B30 |
| url | https://arxiv.org/abs/2605.08362 |